Bioluminescence Monitoring of Neuronal Activity in Freely Moving Zebrafish Larvae

The proof of concept for bioluminescence monitoring of neural activity in zebrafish with the genetically encoded calcium indicator GFP-aequorin has been previously described (Naumann et al., 2010) but challenges remain. First, bioluminescence signals originating from a single muscle fiber can constitute a major pitfall. Second, bioluminescence signals emanating from neurons only are very small. To improve signals while verifying specificity, we provide an optimized 4 steps protocol achieving: 1) selective expression of a zebrafish codon-optimized GFP-aequorin, 2) efficient soaking of larvae in GFP-aequorin substrate coelenterazine, 3) bioluminescence monitoring of neural activity from motor neurons in free-tailed moving animals performing acoustic escapes and 4) verification of the absence of muscle expression using immunohistochemistry.


Background
Unlike fluorescent genetically encoded calcium indicators (GECIs) (Grienberger and Konnerth, 2012), such as the GCaMP family, the bioluminescent indicator GFP-aequorin (Shimomura et al., 1962) does not require light excitation and therefore opens new avenues for monitoring neural activity in moving animals, including flies (Martin et al., 2007), mice  and zebrafish larvae (Naumann et al., 2010). However, efficient use of GFP-aequorin remains challenging to achieve in zebrafish larvae, restricting its widespread use as a calcium indicator. The limitation lies in the fact that bioluminescence signals This article is distributed under the terms of the Creative Commons Attribution License (CC BY 4.0). originating from a single muscle fiber are so large they constitute a major pitfall. Once absence of muscle expression is verified for a given transgenic line, bioluminescence signals emanating from neurons only are very small. To overcome these limitations, we developed a codon-optimized variant of GFP-aequorin for zebrafish larvae, achieved selective expression in motor and sensory neurons using existing transgenic lines, modified coelenterazine soaking protocol in order to conduct experiments at 4 days post-fertilization, created a behavioral bioluminescence assay for monitoring neuronal activity during acoustic evoked stereotyped escape responses in zebrafish larvae.

4.
Maintain adult AB and Tüpfel long fin (TL) strains of Danio rerio on a 14/10 h light cycle and water is maintained at 28.5 °C, conductivity at 500 μS and pH at 7.4.

5.
Raise embryos in 'blue water' (3 g of Instant Ocean ® salts and 2 ml of methylene blue at 1% in 10 L of osmosed water, see Recipes) at 28.5 °C during the first 24 h before screening for GFP expression.

B.
Characterization of GFP-aequorin expression with immunohistochemistry

1.
Fix 4 dpf larvae in 4% PFA for 4 h at 4 °C followed by 3 x 5 min washes in PBS.

2.
Block larvae for 1 h in blocking solution (see Recipes) (agitation required).

4.
Wash three times for 5 min in washing solution (see Recipes), then incubate larvae in the dark with the secondary antibody (Alexa Fluor 488 goat anti-chicken IgG, dilution1:1,000) in PBST (agitation required) for 2 h at RT.

5.
Wash three times for 5 min in PBST, then mount larvae on a slide with mounting medium and image on a standard upright confocal microscope (Olympus FV-1000).

6.
Perform negative IHC controls by omitting the primary antibody.

7.
Image the entire immunostained Tg(mnx1:gal4;UAS:GFP-aequorinopt) icm09 larvae to confirm selective expression of GFP-aequorin-opt in spinal motor neuron populations and absence from muscle fibers. We noted more prominently primary dorsal motor neurons but also  (Figure 1) without any expression in the muscles and only very limited expression in the brain and hindbrain.

C.
Soaking of larvae in coelenterazine solution

5.
Renew 60 μM soaking solution at 2 days post-fertilization. Embryos are maintained in the dark.

6.
Perform behavioral experiments at 4 days post-fertilization (total soaking time is 72 h).

D.
Monitoring neuronal activity with bioluminescence
Using black boards, create a 1 m square lightproof box.

b.
Infrared light illumination is provided by an 850 nm LED mounted with 2 long-pass 780 and 810 filters and a diffuser.

c.
Video acquisition is performed at 1,000 Hz using a high-speed infrared sensitive camera at 320 x 320 pixels resolution controlled by the video software (Hiris ® ).

d.
Photons are counted with a photomultiplier tube located under the larva arena and sent to an acquisition card. A band-pass filter (525 nm/50 nm) and a short-pass filter (670 nm) are placed between the larva and the PMT.

e.
A custom application-programming interface synchronizes the video acquisition with the photon count and the stimulus delivery using a 30 trials batched TTL chronogram.

2.
Run the bioluminescence assay one larva at a time a.
Place larva in a circular (2 cm diameter) 3D-printed arena (larva can also be head-embedded in 1.5% low-melting point agarose with the tail free to move).

b.
Place the larva in the arena and attach the arena to a small 2-Ohm speaker.

c.
Deliver sinusoidal stimuli (5 cycles, 500 Hz) produced by the waveform generator and audio amplifier through a 2-Ohm speaker attached to the larva arena.

d.
Adjust intensity to the lowest value reliably eliciting an escape response (between 0.5 and 5 V usually).

e.
Each trial consists in a 500 msec baseline followed by a 10 msec acoustic stimulus and 1,990 msec subsequent recording.

f.
Assays consist of 30 trials with 1-min inter-trial intervals to reduce habituation.